NYT Lawsuit Names Anthropic, Mistral, and xAI as Downstream Scraped-Data Consumers
What happened
The New York Times and three co-plaintiff publishers filed a copyright lawsuit against Brave Software, reported by The Verge. They allege Brave unlawfully scraped their content and sold access to that index to AI companies. The complaint names Anthropic, Perplexity, Mistral, xAI, and Cohere as downstream purchasers of that allegedly infringing data. A separate charge accuses Brave of enabling users to access publisher content behind paywalls without payment or consent. The case builds on a broader pattern of copyright enforcement against AI data sourcing. Prior actions include Anthropic's $1.5B copyright settlement and earlier reporting that internal emails confirm OpenAI and Microsoft knew scraping was legally indefensible. Compliance teams should note that liability runs through an intermediary. The publishers are not suing the AI companies directly, but the named vendors are central to the alleged harm.
Why it matters
- ·Enterprises using Anthropic, Mistral, xAI, Perplexity, or Cohere should check whether their vendor contracts include indemnification for claims arising from training or retrieval data sourced through third parties. If they do not, the enterprise may carry residual exposure if litigation affects the vendor's data licensing obligations.
- ·This case exposes a gap in standard AI vendor due diligence. Most programs review a tier-1 vendor's own policies but do not ask how that vendor sources training or retrieval index data from intermediaries. A finding that a named vendor paid for access to unlicensed content could affect that vendor's product, pricing, or legal standing.
- ·Content-sensitive sectors, including media, legal services, and financial research, face heightened risk when AI outputs draw on retrieval pipelines that may include scraped content. If a vendor's index is later found to have been improperly sourced, outputs derived from that index could carry copyright taint that affects enterprise use of those outputs.
Stay ahead of stories like this
Get every US AI governance development like this one, plus the rest of the week's developments. Every Thursday.
Recent issues
- Three separate incidents this week each created a new compliance action item: verify your AI vendors' data sourcing, demand agent monitoring coverage guarantees, and audit trusted app identities in your cloud environment.29 Oct
- Regulators, litigants, and attackers are all moving faster than most organizations' AI governance frameworks, shifting leverage away from deployers and toward those who can prove control, accountability, and honest disclosure.21 Oct
Free every Thursday. Unsubscribe anytime.
Governance controls affected
What to do now
- ☐Ask each named vendor (Anthropic, Perplexity, Mistral, xAI, Cohere) whether their training or retrieval index data includes content sourced from Brave or similar web-index intermediaries, and request a written response.
- ☐Review your AI vendor contracts for indemnification clauses that cover third-party intellectual property claims arising from the vendor's data sourcing practices, not just the vendor's direct actions.
- ☐Add a question to your vendor due diligence intake asking specifically how the vendor sources data for training and retrieval pipelines, including whether it uses index or data-broker intermediaries.
- ☐Flag this lawsuit to your legal team and ask them to assess whether outputs your organization has already produced using the named vendors carry any copyright exposure under current case law.
- ☐Monitor the litigation for any court order requiring the named vendors to disclose their data sourcing arrangements, which could affect your own vendor risk assessments.
What to watch next
Courts have not yet ruled on whether downstream AI vendors bear liability for purchasing access to an allegedly infringing data index. That question will shape how far copyright risk travels through the AI supply chain. Compliance teams should watch for any preliminary injunction or discovery order that compels the named vendors to disclose their data sourcing practices. Congress has received copyright litigation referrals before. Most recently, a federal court dismissed Google AI Overviews suits and pointed publishers to Congress. Legislative action on AI training data rights remains a live possibility. The outcome of the NYT v. OpenAI litigation, already underway, will also set precedents that courts handling this Brave action are likely to follow.